Photonic neural network

A method comprising:

• modelling a photonic neural network device, wherein the photonic neural network device comprises:
• a planar waveguide;
• a layer having a changeable refractive index adjacent to the planar waveguide;
• and
• a plurality of electrodes, wherein each electrode is electrically coupled to the layer having the changeable refractive index at a corresponding location of the layer having the changeable refractive index, wherein each electrode is configured to apply a corresponding, configurable voltage to the corresponding location to affect a refractive index of the corresponding location of the layer having the changeable refractive index to configure a corresponding neuron of the photonic neural network device in order to perform a computation;
• training the modelled photonic neural network, wherein training comprises selecting the corresponding, configurable voltages at corresponding locations so that the corresponding neurons produce a desired output waveform in response to an input waveform;
• and
• physically implementing the trained photonic neural network in a device based on the training.

A photonic neural network device comprising:

• a plurality of neural network layers for a neural network, wherein each neural network layer comprises:
• a column of neurons, each neuron comprising a corresponding electrode adjacent to a layer having a changeable refractive index, wherein the layer having the changeable refractive index is adjacent to a planar optical waveguide;
• and
• a column of refractive or diffractive elements adjacent to the planar optical waveguide, wherein the photonic neural network is configured to induce an amplitude or a phase modulation of a light waveform propagating through the photonic neural network device to configure a corresponding neuron of the photonic neural network device in order to perform a computation.